Deceptive Games

Damien Anderson, Matthew Stephenson, Julian Togelius, Christoph Salge, John Levine, Jochen Renz

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    Deceptive games are games where the reward structure or other aspects of the game are designed to lead the agent away from a globally optimal policy. While many games are already deceptive to some extent, we designed a series of games in the Video Game Description Language (VGDL) implementing specific types of deception, classified by the cognitive biases they exploit. VGDL games can be run in the General Video Game Artificial Intelligence (GVGAI) Framework, making it possible to test a variety of existing AI agents that have been submitted to the GVGAI Competition on these deceptive games. Our results show that all tested agents are vulnerable to several kinds of deception, but that different agents have different weaknesses. This suggests that we can use deception to understand the capabilities of a game-playing algorithm, and game-playing algorithms to characterize the deception displayed by a game.

    Original languageEnglish (US)
    Title of host publicationApplications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings
    PublisherSpringer-Verlag
    Pages376-391
    Number of pages16
    ISBN (Print)9783319775371
    DOIs
    StatePublished - Jan 1 2018
    Event21st International Conference on Applications of Evolutionary Computation, EvoApplications 2018 - parma, Italy
    Duration: Apr 4 2018Apr 6 2018

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume10784 LNCS
    ISSN (Print)0302-9743
    ISSN (Electronic)1611-3349

    Other

    Other21st International Conference on Applications of Evolutionary Computation, EvoApplications 2018
    CountryItaly
    Cityparma
    Period4/4/184/6/18

    Fingerprint

    Game
    Deception
    Video Games
    Artificial intelligence
    Artificial Intelligence
    Optimal Policy
    Reward
    Series

    Keywords

    • Deception
    • Games
    • Reinforcement learning
    • Tree search

    ASJC Scopus subject areas

    • Theoretical Computer Science
    • Computer Science(all)

    Cite this

    Anderson, D., Stephenson, M., Togelius, J., Salge, C., Levine, J., & Renz, J. (2018). Deceptive Games. In Applications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings (pp. 376-391). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10784 LNCS). Springer-Verlag. https://doi.org/10.1007/978-3-319-77538-8_26

    Deceptive Games. / Anderson, Damien; Stephenson, Matthew; Togelius, Julian; Salge, Christoph; Levine, John; Renz, Jochen.

    Applications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings. Springer-Verlag, 2018. p. 376-391 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10784 LNCS).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Anderson, D, Stephenson, M, Togelius, J, Salge, C, Levine, J & Renz, J 2018, Deceptive Games. in Applications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10784 LNCS, Springer-Verlag, pp. 376-391, 21st International Conference on Applications of Evolutionary Computation, EvoApplications 2018, parma, Italy, 4/4/18. https://doi.org/10.1007/978-3-319-77538-8_26
    Anderson D, Stephenson M, Togelius J, Salge C, Levine J, Renz J. Deceptive Games. In Applications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings. Springer-Verlag. 2018. p. 376-391. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-77538-8_26
    Anderson, Damien ; Stephenson, Matthew ; Togelius, Julian ; Salge, Christoph ; Levine, John ; Renz, Jochen. / Deceptive Games. Applications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings. Springer-Verlag, 2018. pp. 376-391 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inproceedings{912a58839c1e4ab4b88c8fba0d57c273,
    title = "Deceptive Games",
    abstract = "Deceptive games are games where the reward structure or other aspects of the game are designed to lead the agent away from a globally optimal policy. While many games are already deceptive to some extent, we designed a series of games in the Video Game Description Language (VGDL) implementing specific types of deception, classified by the cognitive biases they exploit. VGDL games can be run in the General Video Game Artificial Intelligence (GVGAI) Framework, making it possible to test a variety of existing AI agents that have been submitted to the GVGAI Competition on these deceptive games. Our results show that all tested agents are vulnerable to several kinds of deception, but that different agents have different weaknesses. This suggests that we can use deception to understand the capabilities of a game-playing algorithm, and game-playing algorithms to characterize the deception displayed by a game.",
    keywords = "Deception, Games, Reinforcement learning, Tree search",
    author = "Damien Anderson and Matthew Stephenson and Julian Togelius and Christoph Salge and John Levine and Jochen Renz",
    year = "2018",
    month = "1",
    day = "1",
    doi = "10.1007/978-3-319-77538-8_26",
    language = "English (US)",
    isbn = "9783319775371",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    publisher = "Springer-Verlag",
    pages = "376--391",
    booktitle = "Applications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings",

    }

    TY - GEN

    T1 - Deceptive Games

    AU - Anderson, Damien

    AU - Stephenson, Matthew

    AU - Togelius, Julian

    AU - Salge, Christoph

    AU - Levine, John

    AU - Renz, Jochen

    PY - 2018/1/1

    Y1 - 2018/1/1

    N2 - Deceptive games are games where the reward structure or other aspects of the game are designed to lead the agent away from a globally optimal policy. While many games are already deceptive to some extent, we designed a series of games in the Video Game Description Language (VGDL) implementing specific types of deception, classified by the cognitive biases they exploit. VGDL games can be run in the General Video Game Artificial Intelligence (GVGAI) Framework, making it possible to test a variety of existing AI agents that have been submitted to the GVGAI Competition on these deceptive games. Our results show that all tested agents are vulnerable to several kinds of deception, but that different agents have different weaknesses. This suggests that we can use deception to understand the capabilities of a game-playing algorithm, and game-playing algorithms to characterize the deception displayed by a game.

    AB - Deceptive games are games where the reward structure or other aspects of the game are designed to lead the agent away from a globally optimal policy. While many games are already deceptive to some extent, we designed a series of games in the Video Game Description Language (VGDL) implementing specific types of deception, classified by the cognitive biases they exploit. VGDL games can be run in the General Video Game Artificial Intelligence (GVGAI) Framework, making it possible to test a variety of existing AI agents that have been submitted to the GVGAI Competition on these deceptive games. Our results show that all tested agents are vulnerable to several kinds of deception, but that different agents have different weaknesses. This suggests that we can use deception to understand the capabilities of a game-playing algorithm, and game-playing algorithms to characterize the deception displayed by a game.

    KW - Deception

    KW - Games

    KW - Reinforcement learning

    KW - Tree search

    UR - http://www.scopus.com/inward/record.url?scp=85044062940&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85044062940&partnerID=8YFLogxK

    U2 - 10.1007/978-3-319-77538-8_26

    DO - 10.1007/978-3-319-77538-8_26

    M3 - Conference contribution

    AN - SCOPUS:85044062940

    SN - 9783319775371

    T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    SP - 376

    EP - 391

    BT - Applications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings

    PB - Springer-Verlag

    ER -